Optic Fiber Basics

8/3/2019 Optic Fiber Basics

1/53

Including

An Introduction

Fiber-Optic Basics

Tables and Terms

Applications

LEARNING ABOUT

OPTIONS IN FIBER

FIBER OPTIONS, INC. / 80 Orville Drive / Bohemia / New York / 11746-2533

5 1 6 -5 6 7 -8 3 2 0 / 1 -8 0 0 -3 4 2 -3 7 4 8 / F AX 5 1 6 -5 6 7 -8 3 2 2


2/53

TABLE OF CONTENTS

SECTION 1

History of Information Transmission .............................................................................................1-1

Advantages/Disadvantages of Fiber............................................................................................1-1

Light and Reflection and Refraction.............................................................................................1-3

SECTION 2

The Optical Fiber ..........................................................................................................................2-1

The Fiber-Optic Cable..................................................................................................................2-7

Sources.......................................................................................................................................2-12

Detectors ....................................................................................................................................2-13

Transmitters and Receivers........................................................................................................2-15

Connectors and Splices .............................................................................................................2-17

Couplers and Networks..............................................................................................................2-25

WDM .......................................................................................................................................2-26

Optical Switch.............................................................................................................................2-27

SECTION 3Tables .........................................................................................................................................3-1

Glossary of Terms.........................................................................................................................3-5

iRev 10/1994


3/53

HISTORY

The use of light for the transmission of information

is far from a new idea. Paul Reveres lanterns wereused to signal the approach of the British. And itwas Alexander Graham Bells experiments over a

century ago that led to his development of thephotophone, a device that carried speech fromone point to another by means of vibrating mirrors

and a beam of sunlight.

Although never a commercial success, it neverthe-

less demonstrated the feasibility of lightwave com-munications. However the technique was shuntedaside and virtually forgotten for almost anotherhundred years.

It probably would have remained in limbo had it notbeen for the appearance of a device called the

laser.

Laser is an acronym for Light Amplification byStimulated Emission of Radiation.

Simply described, the laser is a device that pro-duces a unique kind of radiation an intensely

bright light which can be focused into a narrowbeam of precise wavelength. The tremendousenergies of the laser stem from the fact that it pro-duces what is called coherent light .

The light that comes from a candle or an incan-descent bulb is called incoherent light. It's made

up of many different, relatively short wavelengths(colors) which together appear white. They aresent out in brief bursts of energy at different timesand in different directions. These incoherent light

waves interfere with each other, thus their energyis weakened, distorted, and diffused.

The laser, on the other hand, emits light wavesthat all have the same wavelength, are in phase,and can be sharply focused to travel in the samedirection over long distances with almost no dis-

persion or loss of power.

Lasers provide radiation at optical and infrared

frequencies. With lasers (and associated elec-tronics) it became possible to perform at opticalfrequencies the electronics functions that engi-

neers were accustomed to performing at conven-tional radio and microwave frequencies. Thuslasers promised the ability to channel signals withvery high information rates along an extremely

narrow path.

INFORMATION TRANSMISSION

Fiber optics is a relatively new technology that

uses rays of light to send information over hair-thinfibers at blinding speeds. These fibers are usedas an alternative to conventional copper wire in a

variety of applications such as those associatedwith security, telecommunications, instrumentationand control, broadcast or audio/visual systems.

Today the ability to transmit huge amounts of infor-mation along slender strands of high-purity glass

optical fiber with the speed of light has revolution-ized communications.

The large signal-carrying capacity of optical fibers

makes it possible to provide not only many more,but much more sophisticated signals than couldever be handled by a comparable amount of

copper wire.

ADVANTAGES/DISADVANTAGES

The advantages of fiber over copper include:

Small Size: A 3/8-inch (12 pair) fiber/cableoperating at 140 mb/s can handle as manyvoice channels as a 3-inch diameter copper(900) twisted-pair cable.

Light Weight: The same fiber-optic cableweighs approximately 132 lbs per kilometer.

The twisted pair cable weighs approximately16,000 lbs.

High Bandwidth: Fiber optics has been band-

width tested at over 4-billion bits per secondover a 100 km (60 miles) distance. Theoreticalrates of 50-billion bits are obtainable.

Low Loss: Current single-mode fibers havelosses as low as .2 dB per km. Multimodelosses are down to 1 dB (at 850 or 1300 nm).

This creates opportunities for longer dis-tances without costly repeaters.

Noise Immunity: Unlike wire systems, whichrequire shielding to prevent electromagneticradiation or pick-up, fiber-optic cable is a

dielectric and is not affected by electromag-netic or radio frequency interference. Thepotential for lower bit error rates can increasecircuit efficiency.

SECTION 1INTRODUCTION TO FIBER

1-1


4/53

Transmission Security: Because the fiber is adielectric the fiber does not radiate electro-

magnetic pulses, radiation, or other energy that

can be detected. This makes the fiber/cabledifficult to find and methods to tap into fibercreate a substantial system signal loss.

No Short Circuits: Since the fiber is glass anddoes not carry electrical current, radiate

energy, or produce heat or sparks, the datais kept within the fiber medium.

Wide Temperature Range: Fibers and cablescan be manufactured to meet temperaturesfrom -40F to +200F. Resistance to tempera-tures of 1,000F have been recorded.

No Spark or Fire Hazard: Fiber optics pro-vides a path for data without transmitting

electrical current. For applications in dan-gerous or explosive environments, fiber pro-

vides a safe transmission medium.

Fewer Repeaters: Few repeaters, if any, arerequired because of increased performanceof light sources and continuing increases in

fiber performance.

Stable Performance: Fiber optics is affectedless by moisture which means less corrosion

and degradation. Therefore, no scheduledmaintenance is required. Fiber also hasgreater temperature stability than copper

systems.

Topology Compatibility: Fiber is suitable tomeet the changing topologies and configura-

tions necessary to meet operation growth andexpansions. Technologies such as wave-length division multiplexing (WDM), optical

multiplexing, and drop and insert technolo-gies are available to upgrade and recon-figure system designs.

Decreasing Costs: Costs are decreasing,larger manufacturing volumes, standardiza-tion of common products, greater repeater

spacing, and proven effectiveness of olderpaid for technologies such as multimode.

Nonobsolescence: Expansion capabilitiesbeyond current technologies using commonfibers and transmission techniques.

Material Availability: Material (silica glass)required for the production of fiber is readily

available in a virtually unending supply.

The few disadvantages of fiber include:

Cost: Individual components, such as con-nectors, light sources, detectors, cable andtest equipment, may be relatively expensive

when compared directly to equivalent itemsin a copper system.

Taps: Drop points must be planned becauseoptical splitters or couplers are much moredifficult to install after the system is in.

Fear of New Technologies: Because the tech-nology is considered to be new, people arereluctant to change and use these methods.

The use of metric and physics is still an unfa-miliar area to may established users.

LIGHT

Light is electromagnetic energy, as are radiowaves, radar, television and radio signals, x-rays,and electronic digital pulses. Electromagneticenergy is radiant energy that travels through free

space at about 300,000/km/s or 186,000 miles/s.

An electromagnetic wave consists of oscillating

electric and magnetic fields at right angles to eachother and to the direction of propagation. Thus, anelectromagnetic wave is usually depicted as asine wave. The main distinction then between dif-

ferent waves lies in their frequency or wavelength.In electronics we customarily talk in terms of fre-quency. In fiber optics, however, light is described

by wavelength. Frequency and wavelength areinversely related.

Electromagnetic energy exists in a continuous

range from subsonic energy through radio waves,microwaves, gamma rays, and beyond. This rangeis known as the electromagnetic spectrum.

It seems to be well understood that glass opticalfiber does not conduct electrons as wire does, or

channel radio-frequency signals as coaxial cabledoes. However, many are unclear about how thelight signals are transmitted and how light acts asa messenger for video, audio, and data over fiber.

1-2



5/53

REFLECTION AND REFRACTION

Optical fiber transmits light by a law of physics

known as the principle of total internal reflection.This principle was discovered by a British scientistnamed John Tyndall in the mid-1800s. He used it

to demonstrate a way to confine light and actuallybend it around corners. His experiments directeda beam of light out through a hole in the side of a

bucket of water. He was able to demonstrate howthe light was confined to the curved stream ofwater, and how the waters changing path redi-

rected the path of light.

Total internalreflection is even more efficient thanmirrored reflection; it reflects more than 99.9percent of the light.

The quantifiable physical property of a transparent

material that relates to total internal reflection is its

refractive index. Refractive index is defined as the

ratio of the speed of light in a vacuum to thespeed of light in a specific material.

Light travels fastest through a vacuum. As it startsto travel through denser material, it slows down a

little. What is commonly called the speed of light isactually the velocity of electromagnetic energy in avacuum such as space. Light travels at slowervelocities in other materials such as glass.

Light traveling from one material to anotherchanges speed, which results in light changing its

direction of travel. This deflection of light is called

refraction. In addition, different wavelengths oflight travel at different speeds in the same mate-rial. The variation of velocity with wavelength plays

an important role in fiber optics.

White light entering a prism contains all colors.

The prism refracts the light and it changes speedas it enters. Because each wave changes speeddifferently, each is refracted differently. Red lightdeviates the least and travels the fastest. Violet

light deviates the most and travels the slowest.

The light emerges from the prism divided into the

colors of the rainbow. As can be seen in Figure 1-1refraction occurs at the entrance and at the exit ofthe prism. The amount that a ray of light is refracted

depends on the refractive indices of the two mate-rials. Figure 1-2 illustrates several important termsrequired to understand light and its refraction.

The normal is an imaginary line perpendicular

to the interface of the two materials.

The angle of incidence is the angle betweenthe incident ray and the normal.

The angle of refraction is the angle betweenthe refracted ray and the normal.

Light passing from a lower refractive index to ahigher one is bent toward the normal. But light

going from a higher index to a lower one refractsaway from the normal, as shown in Figure 1-3.

1-3

Refraction

Refraction Red

Orange

Yellow

Green

Blue

Violet

InterfaceAngle of Refraction

Refracted Ray

Reflected

Wave

Normal

Incident

Ray

Angle of

Incidence

n1n

2

is less than n2n1


Figure 1-1Refraction and a Prism

Figure 1-2Angles of Incidence and Refraction


6/53

As the angle of incidence increases, the angle ofrefraction of 90 is the critical angle. If the angle of

incidence increases past the critical angle, thelight is totally reflected back into the first materialso that it doesnt enter the second material. The

angles of incidence and reflection are equal.

Thus:

Light is electromagnetic energy with a higherfrequency and shorter wavelength than radiowaves.

Light has both wave-like and particle-like char-acteristics.

When light meets a boundary separatingmaterials of different refractive indices, it iseither refracted or reflected.

1-4

When the angle of incidence is morethan the critical, light is reflected

n1n2

Angle ofIncidence

Angle ofReflection

=

Angle ofRefraction

Light is bent away from normal

Angle ofIncidence

n1n2

Light does not enter second material

CritialAngle

n1n2

is greater than n2n1


Figure 1-3Refraction


7/53

THE OPTICAL FIBER

BASIC FIBER CONSTRUCTION

Optical fiber consists of a thin strand (or core) ofoptically pure glass surrounded by another layer

of less pure glass (the cladding). The inner coreis the light-carrying part. The surroundingcladding provides the difference in refractive

index that allows total internal reflection of lightthrough the core. The index of the cladding is lessthan 1 percent lower than that of the core.

Most fibers have an additional coating around thecladding. This coating, usually one or more layers ofpolymer, protects the core and cladding from

shocks that might affect their optical or physicalproperties. The coating has no optical propertiesaffecting the propagation of light within the fiber.

Thus the buffer coating serves as a shock absorber.

Figure 2-1 shows the idea of light traveling througha fiber. Light injected into the fiber and striking the

core-to-cladding interface at greater than the crit-ical angle reflects back into the core. Since theangles of incidence and reflection are equal, the

reflected light will again be reflected. The light willcontinue zig zagging down the length of the fiber.

Light, however, striking the interface at less than

the critical angle passes into the cladding where itis lost over distance. The cladding is usually ineffi-cient as a light carrier, and light in the cladding

becomes attenuated fairly rapidly.

Notice also in Figure 2-1 that the light is refractedas it passes from air into the fiber. Thereafter, its

propagation is governed by the indices of the coreand cladding (and by Snells law.) Refer to section

3, Glossary of Terms, for a definition of Snells Law.

The specific characteristics of light propagationthrough a fiber depends on many factors including:

The size of the fiber; the composition of the fiber;and the light injected into this fiber. An under-standing of the interplay between these properties

will clarify many aspects of fiber optics.

Fiber is basically classified into three groups:

Glass (silica) which includes single-mode stepindex fibers, multimode graded index, andmultimode step index.

Plastic clad silica (PCS).

Plastic.

Most optical fibers for telecommunications aremade 99 percent of silica glass, the material from

which quartz and sand are formed. Figure 2-1 onthe previous page shows a fiber, which consists ofan inner core (about 8 to 100 micrometers, or

0.0003 to 0.004 inches, in diameter), a cladding(125 to 140 micrometers outer diameter) and abuffer jacket for protection.

The clad is made of glass of a slightly differentformula. This causes light entering the core at oneend of the fiber to be trapped inside, a phenom-

enon called internal reflection. The light hits theboundary between the core and the claddingbouncing off the cladding much like a billiard balland at the same angle as it travels down the fiber.

2-1

Jacket

Cladding

Core

Cladding

Jacket

Angle of

Incidence

Angle of

Reflection

Light is propagated by

total internal reflection

Light at less than critical

angle is absorbed in jacket

Jacket

Core

Cladding

SECTION 2FIBER-OPTIC BASICS

Figure 2-1Internal Reflection in an Optical Fiber


8/53

Plastic fibers are much larger in diameter and canonly be used for slow-speed, short-distance trans-

mission. Plastic-clad silica (PCS) fibers, featuring a

glass core with a plastic cladding, come betweenglass and plastic fibers in size and performance.Plastic and PCS fibers cost less than silica glass

fibers, but they are also less efficient at transmittinglight. For this reason, they are being used in cars,sensors, and short-distance data-communications

applications.

There are other types of fiber emerging on the

marketplace, particularly suited for specializeduses. An example would be fluoride fibers whichare being developed for medical and long-haultelecommunications. Medical applications for fiber

include transmitting power from a laser to destroyarterial blockages or cancer masses. Since fibersare extremely narrow and flexible, they can be

threaded through human arteries to locate precisetrouble areas, and in some cases may eliminatethe need for surgery.

MODE

James Clerk Maxwell, a Scottish physicist in the

last century, first gave mathematical expression tothe relationship between electric and magneticenergy. Mode is a mathematical and physicalconcept describing the propagation of electro-

magnetic waves through media. In its mathemat-ical form, mode theory derives from Maxwellsequations. He showed that they were both a single

form of electromagnetic energy, not two differentforms as was then believed. His equations alsoshowed that the propagation of this energy fol-lowed strict rules.

A mode is simply a path that a light ray can followin traveling down a fiber. The number of modes

supported by a fiber ranges from one to over100,000. Thus a fiber provides a path of travels for

one or thousands of light rays, depending on its

size and properties.

REFRACTIVE INDEX PROFILE

This term describes the relationship between theindices of the core and the cladding. Two main

relationships exist: Step index and graded index.The step-index fiber has a core with a uniformindex throughout. The profile shows a sharp step

at the junction of the core and cladding. In contrast,the graded index has a nonuniform core. The indexis highest at the center and gradually decreasesuntil it matches that of the cladding. There is no

sharp break between the core and the cladding.

Step Index

The multimode step-index fiber is the simplesttype. It has a core diameter from 100 to 970 m,and it includes glass, PCS, and plastic construc-

tions. As such, the step-index fiber is the mostwide ranging, although not the most efficient inhaving high bandwidth and low losses.

Graded Index

A graded-index fiber is one where the refractive

index of the fiber decreases radically towards theoutside of the core. During the manufacturingprocess, multiple layers of glass are deposited on

the preform in a method where the optical indexchange occurs. (Refer Figure 2-3 next page.)

As the light ray travels through the core, the

fastest index is the higher or outer area in agraded-index core. (Refer Figure 2-4 next page.)

2-2

Silica Glass Core HighRefractive Index

Plastic JacketPlastic Cladding

Low Refractive Index

Light

SECTION 2FIBER OPTIC BASICS

Figure 2-2Plastic-Clad Silica Fiber


9/53

The center, or axial mode would be the slowest modein the graded-index fiber Figure 2-5). In this circum-

stance, a mode would slow down when passing

through the center of the fiber and accelerate whenpassing through the outer areas of the core. This isdesigned to allow the higher order modes to arrive

at approximately the same time as an axial or lowerorder mode. This allows the multimode graded-index fibers to transmit as far as 15-20 kilometers

without great pulse spreading. Within these classifi-cations there are three types of fiber:

Multimode step-index. Multimode graded-index. Single-mode step-index.

STEP INDEX

Multimode Step-Index Fiber

Bandwidth of 10 MHz/km

Loss of 5-20 dB/km. Large cores of 200 to 1000 microns.

Cladding OD up to 1035 microns.

Is effective with low-cost LEDs

Limited transmission distances.

Transmits at 660-1060 wavelengths.

Single-Mode Step Index Fiber

High bandwidth applications (4 GHz).

Low losses, typically .3 dB to .5 dB/km.

Core area of 8 to 10 microns.

Cladding OD of 125 microns.

Transmits at 1300 nm and 1550 nm wave-lengths.

Higher costs for connectors, splices, and testequipment, and transmitters/receivers.

Plastic Step-Index Fiber

Lower bandwidth 5 MHz over distances of200 feet.

Losses of 150-250 dB/km.

Core area from 1000-3000 microns.

Cladding up to 3000 microns.

Uses LEDs to transmit data very well.

Very easy to connectorize.

Inexpensive.

Operates best at 660 nm red wavelength.

Plastic-Clad Silica Step-Index Fiber

Bandwidth up to 25 MHz/km

Losses of 6-10 dB/km.

Glass core from 200-600 microns.

Plastic cladding OD to 1000 microns.

LEDs used to transmit data.Difficult to connectorize and unstable.

Very resistant to radiation.

Operates at 660-1060 wavelengths.

GRADED INDEX

Multimode Graded-Index Fiber

Bandwidths up to 600 MHz/km.

Losses of 2 to 10 dB/km.

Cores of 50/62.5/85/100 microns.

Cladding OD of 125 and 140 microns.

Is effective with laser or LED sources.

Medium- to low-cost for components, test

equipment, and transmitters and receivers.2-3

Light rays passing throughmultiple layers of glass

n1

n6

n5

n4

n3

n2


Figure 2-3Graded Index Fiber

Figure 2-4High-Order Mode

Figure 2-5Low-Order Mode


10/53

Has distance limitations due to higher lossand pulse spreading.

Transmits at 820-850 nm, 1300 nm, and 1550nm wavelengths.

Easy to splice and connectorize.

MULTIMODE AND SINGLE-MODE FIBER

Two general types of fiber have emerged to meetuser requirements: multimodeand single mode.In

optical terminology, mode can be thought of asa ray of light.

In multimode fiber many modes, or rays, are trans-

mitted, whereas in single-mode fiber only onemode of light can travel in the core. Refer to Figure2-6 where the core diameters of these two types of

fiber have been compared to the diameter of asingle human hair.

Multimode

Multimode fibers larger core (diameter in the

50-m to more than 1000-m range) captures hun-dreds of rays from the light source, entering thecore at many different angles. Some of these raysexceed the critical angle of incidence and are lost

without penetrating the fiber.

Of the rays that are captured by the core, some

travel a direct path parallel to the length of thefiber. Modes that enter at a steeper angle travel alonger, circuitous route, crisscrossing the coresdiameter as they travel down the fiber. Because of

these different routes, some parts of the light pulsereach the far end sooner than other parts of thesame light pulse.

These differences result in pulse broadening (orspreading) which requires more space betweenpulses, thereby limiting the speed at which pulses

can be introduced into the fiber, and limiting thebandwidth or information-carrying capacity of mul-timode fiber.

Multimode fibers were developed first, and they

have been installed in many long-distance tele-communications systems. In the past few years,however, single-mode technology has improved tothe point where these smaller fibers are made aseasily and as cheaply as multimode fibers.

Multimode fibers significantly larger core (more thanfive times the diameter of a single-mode core) hascertain advantages. It is easier to align core regions

for splicing and for attaching connectors, and it cap-tures more light from lower cost sources, such asfrom LEDs rather than lasers. Thus multimode isusually preferred for systems that have many con-

nectors or joints, and where distance or capacity isnot a factor.

Further, methods can be devised for increasingmultimode fibers information-carrying capacity,such as transmitting on multiple wavelengths oflight. This technique is known as wavelength divi-

sion multiplexing or WDM.

Single-Mode

Single-mode fiber overcomes the bandwidth short-

comings of multimode. Single-mode fiber has amuch smaller core diameter (typically 8 m to 10m) allowing a very narrow beam from a singlesource to pass through it with a minimum of pulsedispersion. The cladding diameter, however,

2-4


Core

8 m

125 m

Cladding

Core50 m

125 m

Cladding

75 m

Figure 2-6Core Diameter of Fiber

Human Hair

Multimode

Single Mode


11/53

remains at the industry standard of 125 micronsfor purposes of connecting and splicing.

With only one mode it is easier to maintain theintegrity of each light pulse. The pulses can bepacked much more closely together in time, giving

single-mode fiber much larger channel capacity.

Refer to section 3, References, Tables A and B for

charts offering fiber comparisons.

DISPERSION

Dispersion is the spreading of a light pulse as ittravels down the length of an optical fiber. Dispersionlimits the bandwidth (or information-carrying capacity)

of a fiber. There are three main types of dispersion:Modal, material, and waveguide.

Modal Dispersion

Modal dispersion occurs only in multimode fiber.

Multimode fiber has a core diameter in the 50-mto more than 1000-m range. The large coreallows many modes of light propagation. Since

light reflects differently for different modes, somerays follow longer paths than others. (Refer topage 2-3, Figures 2-3, 2-4 and 2-5.)

The lowest order mode, the axial ray travelingdown the center of the fiber without reflecting,arrives at the end of the fiber before the higher

order modes that strike the core-to-cladding inter-face at close to the critical angle and, therefore,follow longer paths.

Thus, a narrow pulse of light spreads out as ittravels through the fiber. This spreading of a lightpulse is called modal dispersion. There are three

ways to limit modal dispersion:

Use single-mode fiber since its core diameteris small enough that the fiber propagates only

one mode efficiently.

Use a graded-index fiber so that the light rays

that follow longer paths also travel at a fasteraverage velocity and thereby arrive at the

other end of the fiber at nearly the same timeas rays that follow shorter paths.

Use a smaller core diameter, which allowsfewer modes.

Material Dispersion

Different wavelengths (colors) also travel at different

velocities through a fiber, even in the same mode(refer to earlier discussions on Index of Refraction).Each wavelength, however, travels at a different

speed through a material, so the index of refrac-tion changes according to wavelength. This phe-nomenon is called material dispersion since it

arises from the material properties of the fiber.

Material dispersion is of greater concern in single-mode systems. In multimode systems, modal dis-

persion is usually significant enough that materialdispersion is not a problem

Waveguide Dispersion

Waveguide dispersion, most significant in a single-

mode fiber, occurs because optical energy travelsat slightly different speeds in the core andcladding. This is because of the slightly differentrefractive indices of the materials.

Altering the internal structure of the fiber allowswaveguide dispersion to be substantially

changed, thus changing the specified overall dis-persion of the fiber.

BANDWIDTH VS. DISPERSION

Manufacturers of multimode offerings frequentlydo not specify dispersion, rather they specify a

measurement called bandwidth (which is given inmegahertz/kilometers).

For example, a bandwidth of 400 MHz/km means

that a 400-MHz signal can be transmitted for 1 km.It also means that the product of the frequencyand the length must be 400 or less (BW x L =

400). In other words, you can send a lower fre-quency a longer distance: 200 MHz for 2 km; 100MHz for 4 km; or 50 MHz for 8 km.

Conversely, a higher frequency can be sent ashorter distance: 600 MHz for 0.66 km; 800 MHzfor 0.50 km; or 1000 MHz for 0.25 km

Single-mode fibers, on the other hand, are speci-

fied by dispersion. This measurement is expressedin picoseconds per kilometer per nanometer ofsource spectral width (ps/km/nm).

In other words, for single-mode fiber dispersion is

2-5



12/53

most affected by the sources spectral width; thewider the source width (the more wavelengths

injected into the fiber), the greater the dispersion.

ATTENUATION

Attenuation is the loss of optical power as lighttravels through fiber. Measured in decibels perkilometer, it ranges from over 300 dB/km for

plastic fibers to around 0.21 dB/km for single-mode fiber.

Attenuation varies with the wavelength of light. In

fiber there are two main causes: Scattering andAbsorption

Scattering

Scattering (Figure 2-7), the more common source

of attenuation in optical fibers, is the loss of opticalenergy due to molecular imperfections or lack ofoptical purity in the fiber and from the basic struc-ture of the fiber.

Scattering, does just what its name implies. It scat-ters the light in all directions including back to the

optical source. This light reflected back is whatallows optical time domain reflectometers (OTDRs)to measure attenuation levels and optical breaks

AbsorptionAbsorption (Figure 2-8) is the process by which

impurities in the fiber absorb optical energy anddissipate it as a small amount of heat, causing thelight to become dimmer. The amount convertedto heat, however, is very minor.

Microbend Loss

Microbend loss (Figure 2-9) results from small varia-tions or bumps in the core-to-cladding interface.Transmission losses increase due to the fiber radiusdecreasing to the point where light rays begin to

pass through the cladding boundary. This causesthe fiber rays to reflect at a different angle, thereforecreating a circumstance where higher order modes

are refracted into the cladding to escape. As theradius decreases, the attenuation increases.

Fibers with a graded index profile are less sensi-tive to microbending than step-index types. Fiberswith larger cores and different wavelengths canexhibit different attenuation values.

Macrobend Loss

Macrobend losses (Figure 2-10) are caused bydeviations of the core as measured from the axis

of the fiber. These irregularities are caused duringthe manufacturing procedures and should not beconfused with microbends.

2-6


Figure 2-7Scattering

Figure 2-8Absorption

Figure 2-9Microbend

Figure 2-10Macrobend


13/53

NUMERICALAPERTURE

The numerical aperture (NA), or light-gathering

ability of a fiber, is the description of the maximumangle in which light will be accepted and propa-gated within the core of the fiber. This angle of

acceptance can vary depending upon the opticalcharacteristics of the indices of refraction of thecore and the cladding.

If a light ray enters the fiber at an angle which isgreater than the NA or critical angle, the ray willnot be reflected back into the core. The ray will

then pass into the cladding becoming a claddingmode, eventually to exit through the fiber surface.The NA of a fiber is important because it gives an

indication of how the fiber accepts and propagateslight. A fiber with a large NA accepts light well; afiber with a low NA requires highly directional light.

Fibers with a large NA allow rays to propagate athigher or greater angles. These rays are calledhigher order modes. Because these modes take

longer to reach the receiver, they decrease thebandwidth capability of the fiber and will havehigher attenuation.

Fibers with a lower NA, therefore, transmit lowerorder modes with greater bandwidth rates and lowerattenuation.

Manufacturers do not normally specify NA for single-mode fibers because NA is not a critical parameter

for the system designer or user. Light in a single-mode fiber is not reflected or refracted, so it doesnot exit the fiber at angles. Similarly, the fiber doesnot accept light rays at angles within the NA and

propagate them by total internal reflection. Thus NA,although it can be defined for a single-mode, isessentially meaningless as a practical characteristic.

FIBER STRENGTH

One expects glass to be brittle. Yet, a fiber can be

looped into tight circles without breaking. It can alsobe tied into loose knots (pulling the knot tight willbreak the fiber). Tensile strength is the ability of a

fiber to be stretched or pulled without breaking.

The tensile strength of a fiber exceeds that of asteel filament of the same size. Further, a copperwire must have twice the diameter to have thesame tensile strength as fiber.

As discussed under "Microbend Loss," the maincause of weakness in a fiber is microscopic cracks

on the surface, or flaws within the fiber. Defects

can grow, eventually causing the fiber to break.

BEND RADIUS

Even though fibers can be wrapped in circles,they have a minimum bend radius. A sharp bend

will snap the glass. Bends have two other effects:

They increase attenuation slightly. Thiseffect should be intuitively clear. Bends

change the angles of incidence andreflection enough that some high-ordermodes are lost (similarly to microbends).

Bends decrease the tensile strength ofthe fiber. If pull is exerted across a bend,

the fiber will fail at a lower tensile strengththan if no bend were present.

FIBER-OPTIC CABLE

CABLE CHARACTERISTICS

Fiber-optic cable is jacketed glass fiber. In orderto be usable in fiber-optic systems, the somewhat

fragile optical fibers are packaged inside cablesfor strength and protection against breakage, aswell as against such environmental hazards as

moisture, abrasion, and high temperatures.

Packaging of fiber in cable also protects the fibersfrom bending at too sharp an angle, which could

result in breakage and a consequent loss of signal.

Multiconductor cable is available for all designs

and can have as many as 144 fibers per cable. Itis noteworthy that a cable containing 144 fiberscan be as small as .75 inches in diameter.

In addition to the superior transmission capabilities,small size, and weight advantages of fiber-opticcables, another advantage is found in the absence

of electromechanical interference. There are nometallic conductors to induce crosstalk into the

system. Power influence is nonaffecting, and secu-rity breaches of communications are (at this time)very difficult due to the complexities of tappingoptical fiber.

2-7



14/53

MAIN PARTS OF AFIBER-OPTIC CABLE

The creation of fiber-optic cables involves placingseveral fibers together in a process that involvesuse of strength members and insulated (buffered)conductors. When a number of optical fibers are

placed into a single cable, they are frequentlytwisted around a central passive support (strengthmember) which serves to strengthen the cable.

Although fiber-optic cable comes in many varieties,most have the following elements in common:

Optical fiber (core and cladding, pluscoating).

Buffer.

Strength member.

Jacket.

Previous sections have dealt with fiber, so only theremaining three items will be dealt with now.

Buffer

Fiber coating, or the buffer, serves three purposes:(1) Protection of the fiber surface from mechanical

damage; (2) isolation of the fiber from the effects ofmicrobends; and (3) as a moisture barrier.

The outer layer, or secondary coating, is the toughmaterial that protects the fiber surface from mech-

anical damage during handling and cabling opera-tions. The inner, or primary coating, is a materialdesigned to isolate the fiber from damage frommicrobending. Both layer obviously serve as mois-ture barriers.

With the exception of abrasion, uncoated fiber is vir-tually unaffected by many environments. Because

of this, most environmental tests are designed toevaluate coating performance over time.

The simplest buffer is the plastic coating applied

by the fiber manufacturer to the cladding. An addi-tional buffer is added by the cable manufacturer.The cable buffer is one of two types: loose bufferor tight buffer.

The tight bufferdesign features one or two layersof protective coating placed over the initial fiber

coating which may be on an individual fiber basis,or in a ribbon structure. The ribbon design typi-cally features 12 fibers placed parallel betweentwo layers of tape with the ribbons lying loosely

within the cable core.

An advantage to the tight buffer is that it is more

flexible than loose and allows tighter turn radii.This can make tight-tube buffers useful for indoorapplications where temperature variations areminimal and the ability to make tight turns inside

walls is a desirable feature.

The loose bufferdesign features fibers placed into

a cavity which is much larger than the fiber with itsinitial coating, such as a buffer tube, envelope, orslotted core. This allows the fiber to be slightlylonger than its confining cavity, and allows move-

ment of the fiber within the cable to relieve strainduring cabling and field-placing operations.

Individual tight-buffered fiber cables are not gen-erally used in applications subjected to tempera-

ture changes due to the added attenuationcaused by the strain that is placed on fiber duringthe cabling process and the contraction differ-ences of the coating material and glass fiberswhen subjected to these changes.

2-8

Black

Polyurethane

Outer Jacket

Buffer Jacket

Silicone Coating

Strength

Members

Cladding (Silica)

Core (Silica)Optical Fiber


Figure 2-11Main Parts for a Fiber-Optic Cable

Loose Buffer

Unbuffered

Optical Fiber

Tight Buffer

Buffer Layers Applied

Directly Over Fiber

Figure 2-12Loose and Tight Buffers


15/53

In loose-buffer tube designs, the fiber tube is usuallyfilled with a viscous gel compound which repels

water. Slotted, or envelope designs are usually filledwith a water-repellent powder. Although water doesnot affect the transmission properties of optical fiber,the formation of ice within the cable will cause

severe microbending and added dB loss to thesystem.

A comparison of loose tube features to tight tube isprovided in section 3, Table C.

Strength MemberStrength members add mechanical strength to thefiber. During and after installation, the strength

members handle the tensile stresses applied tothe cable so that the fiber is not damaged.

The most common strength members are of Kevlararamid yarn, steel, and fiberglass epoxy rods.Kevlar is most commonly used when individualfibers are placed within their own jackets. Steel

and fiberglass members are frequently used inmultifiber cables.

Jacket

The jacket, like wire insulation, provides protectionfrom the effects of abrasion, oil, ozone, acids,

alkali, solvents, and so forth. The choice of the

jacket material depends on the degree of resis-tance required for different influences and on cost.

A comparison of the relative properties of variouspopular jacket materials is provided in section 3,Table D.

ADDITIONAL CABLE CHARACTERISTICS

Cables come reeled in various lengths, typically1 or 2 km, although lengths of 5 or 6 km are avail-able for single-mode fibers. Long lengths aredesirable for long-distance applications since

cables must be spliced end-to-end over the lengthof the run, hence the longer the cable, the fewerthe splices that will be required.

Fiber coatings or buffer tubes or both are oftencoded to make identification of each fiber easier.In the long-distance link its necessary to be able

to ensure that fiber A in the first cable is spliced tofiber A in the second cable, and fiber B to fiber B,and so on.

In addition to knowing the maximum tensile loadsthat can be applied to a cable, it's necessary to

know the installation load. This is the short-termload that the fiber can withstand during the actualprocess of installation. This figure includes theadditional load that is exerted by pulling the fiber

through ducts or conduits, around corners, etc.The maximum specified installation load will estab-lish the limits on the length of the cable that can

be installed at one time, given the particular appli-cation.

The second load specified is the operating load.

During its installed life, the cable cannot withstand

loads as heavy as it withstood during installation.The specified operating load is therefore less than

the installation load. The operating load is alsocalled the static load. For the purposes of this dis-cussion we have divided the discussion on cablesby indooror outdoor.

2-9


Simplex

KevlarStrengthMember Outer

Jacket

0.9[.035]

Buffered OpticalFiber

3.0[.118]

KevlarStrengthMember

Duplex

0.9[.035]

OuterJacket

KevlarStrengthMember

BufferedOptical Fiber

3.0[.118]

6.1[.240]

Duplex

0.9[.035]

OuterJacket

BufferedOptical Fiber

5.6[.220]

3.1[.122]

Figure 2-13Indoor Cables


16/53

Indoor Cable

Cables for indoor applications see Figure 2-13below) include:

Simplex

Duplex

Multifiber

Undercarpet

Heavy- and light-duty

Plenum

Simplexis a term used to indicate a single fiber.Duplexrefers to two optical fibers. One fiber maycarry the signals in one direction; the other fiber may

carry the signals in the opposite direction. (Duplexoperation is possible with two simplex cables.)

Physically, duplex cables resemble two simplexcables whose jackets have been bonded together,similar to the jacket of common lamp cords. Thistype of cable is used instead of two simplex

cables for aesthetic reasons and for convenience.Its easier to handle, theres less chance of the twochannels becoming confused, and the appear-

ance is more pleasing.

Multifibercable, as the name would imply, containmore than two fibers. They allow signals to be dis-

tributed throughout a building. Multifiber cables

often contain several loose-buffer tubes, each con-taining one or more fibers. The use of several tubes

allows identification of fibers by tube, since bothtubes and fibers can be color coded.

Undercarpetcable,as this name implies, is runacross a floor under carpeting. It is frequentlyfound in open-space office or work areas that aredefined by movable walls, partitions. A key feature

of this cable is its ability to be rearranged or

reconfigured as space needs change. Oneproblem, however, is making turns without

stressing the fibers. Unfortunately, the fiber on theoutside of the turn must always take a longer paththan the fiber on the inside. This unequal pathlength places differing stresses on the fibers.

(Refer to Figure 2-14 below.)

Heavy- and light-dutycables refer to the rugged-ness of the cable, one being able to withstandrougher handling than the other, especially duringinstallation.

A plenumis the return or air-handling space locatedbetween a roof and a dropped ceiling. The NationalElectrical Code (NEC) has designated strict require-

ments for cables used in these areas.

Because certain jacket materials give off noxious

fumes when burned, the NEC states that cables runin plenum must either be enclosed in fireproof con-duits or be insulated and jacketed with low-smokeand fire-retardant materials.

Thus plenum cables are those whose materialsallow them to be used without conduit. Because

no conduit is used for these cables, they are easierto route. So, while plenum cables initially are moreexpensive, there are savings inherent in installation.

Other benefits are reduced weights on ceilings or fix-

tures and easier reconfigurations and flexibility forlocal area networks and computer data systems.

Outdoor Cable

Cables for outdoor applications include:

Aerial or overhead (as found strung between

buildings or telephone poles).

Direct burial cables that are placed directly in

2-10


Black ThermoplasticJacket

1.91 [.075]

29.4 [1.16]

Cable Strength Member

Optical Fiber

Figure 2-14Undercarpet Cable


17/53

a trench dug in the ground and then covered.

Indirect burial, similar to direct burial, but the

cable is inside a duct or conduit.

Submarine cable is underwater, includingtransoceanic application.

All of the foregoing cables must be rugged anddurable since their applications subject them to avariety of extremes. Typically, the internal glass

fiber is the same for all types of fiber cable withsome small exceptions.

Cables designed for underground use may contain

one or more fibers encased in metal jackets andflooded with a moisture-proofing gel.

Section 3, Table E,offers a chart of ques-tions that should be

addressed whenselecting cables forvarious requirements.

Hybrid Cable

This is a unique type of

cable generally avail-able on special orderonly. It is designed formultipurpose applica-

tions where both opti-

cal fiber and twistedpair wires are jacketed

together in those situa-tions where both tech-nologies are called for.This style cable is also

useful when futureexpansion plans callfor optical fiber.

Hybrid cable (Figure 2-15) allows for existingcopper networks to be

upgraded to fiberwithout the require-ment for new cable.

With hybrid cable, thiscan be accomplished

without disrupting the existing service.

This cable style is also useful in applications suchas local area networks (LANs) and integrateddigital services networks (ISDNs) where easy or

smooth transition from copper to fiber is possibleat a future time, basically because the hybrid

cable permits the end user to be fiber ready.

Cable designs are available with multiple elementsincluding the specific wire or fiber types (single- or

multimode). Fibers are color coded for ready iden-tification. As with conventional cable, hybrids canbe manufactured to specific requirements.

Breakout Cable

A breakout cable is one which offers a rugged

cable design for shorter network designs. Thismay include LANs, data communications, videosystems, and process control environments.

A tight buffer design isused along with indivi-

dual strength membersfor each fiber. Thispermits direct termina-

tion to the cablewithout using breakoutkits or splice panels.Due to the increased

strength of Kevlarmembers, cables areusually heavier andphysically larger than

the telecom types with

equal fiber counts.

The term breakoutdefines the key pur-pose of the cable. Thatis, one can break out

several fibers at anylocation, routing otherfibers elsewhere. For

this reason breakoutcables are, or shouldbe, coded for ease ofidentification.

Because this type ofcable is found in many

building environmentswhere codes may

require plenum cables, most breakout cables meetthe NEC's requirements. The cable is available in avariety of designs that will accommodate the topologyrequirements found in rugged environments. Fibercounts from simplex to 256 are available.

2-11


Figure 2-15Hybrid Cable


18/53

CABLE SELECTION

The design and materials used in the cable con-struction selected will depend upon the environ-ment and operation of the users application. Thevariables are numerous and they will all have to be

carefully weighed.

Refer to section 3, Table E, for a check-off sheet

which may be copied or adapted for for use whensetting out to determine precisely which cable isbest suited for individual applications. This chartshows many, if not all, of the variables that will

have to be considered throughout this process.

SOURCES

At each end of a fiber-optic link is a device for

converting energy from one form to another. At thesource is an electro-optic transducer, which con-verts an electrical signal to an optical signal. At

the other end is the optoelectronic transducerwhich converts optical energy to electrical energy.This is discussed further on the next page underDetectors.

SEMICONDUCTOR PNJUNCTION

The semiconductor pn junction is the basic struc-

ture used in the electro-optic devices for fiber

optics. Lasers, LEDs, and photodiodes all use thepn junction, as do other semiconductor devices

such as diodes and transistors.

LASERS AND LEDS

Optical signals begin at the source with lasers orLEDs transmitting light at the exact wavelength atwhich the fiber will carry it most efficiently. The

source must be switched on and off rapidly andaccurately enough to properly transmit the signals.

Lasers are more powerful and operate at faster

speeds than LEDs, and they can also transmitlight farther with fewer errors.

LEDs, on the other hand, are less expensive, morereliable, and easier to use than lasers. Lasers are

primarily used in long-distance, high-speed trans-mission systems, but LEDs are fast enough andpowerful enough for short-distance communica-tions, including video communications.

Lasers and LEDs are both semiconductor devicesthat come in the form of tiny chips packaged in

either TO-style cans that plug into printed circuitboard or microlens packages, which focus thebeam into the fiber.

LEDs used in fiber optics are made of materialsthat influence the wavelengths of light that areemitted. LEDs emitting in the window of 820 to 870

nm are usually gallium aluminum arsenide(GaAIAs).

Window, in this usage, is a term referring to

ranges of wavelengths matched to the propertiesof the optical fiber. Long wavelength devices foruse at 1300 nm are made of gallium indium

arsenide phosphate (GaInAsP), as well as othercombinations of materials.

Lasers provide stimulated emission rather than thesimplex spontaneous emission of LEDs. The maindifference between a LED and a laser is that the

laser has an optical cavity required for lasing. Thiscavity is formed by cleaving the opposite end ofthe chip to form highly parallel, reflective, mirror-like finishes.

Laser light has the following attributes:

Nearly monochromatic: The light emitted

has a narrow band of wavelengths. It is

nearly monochromaticthat is, of a singlewavelength. In contrast to the LED, laser

light is not continuous across the band of itsspecial width. Several distinct wavelengths areemitted on either side of the centralwavelength.

Coherent: The light wavelengths are inphase, rising and falling through the sine-wave

cycle at the same time.

Highly directional: The light is emitted in ahighly direction pattern with little diver-

gence. Divergence is the spreading of alight beam as it travels from its source.

SOURCE CHARACTERISTICS

Refer to section 3, Table F, for a comparison of themain characteristics of interest for both LED andlaser sources.

2-12



19/53

SPECTRALWIDTH

Earlier, we discussed material dispersion and thefact that different wavelengths travel through afiber at different velocities. The dispersionresulting from different velocities of different wave-

lengths limits bandwidth.

Lasers and LEDs do not emit a single wavelength;

they emit a range of wavelengths. This range isknown as the spectral width of the source. It ismeasured at 50 percent of the maximum ampli-tude of the peak wavelength.

DETECTORS

The detector in the fiber-optic system converts theoptical signal into an electrical signal compatible

with conventional equipment and communicationsnetworks.

A good signal detector responds well to light atthe peak intensity wavelength of the light sourceand fiber combination used (800-900 nanometers,1,000-2,000 nanometers). It also operates with low

interference, has high reliability, long operatinglife, and small size.

PHOTODIODE BASICS

In moving from the conduction band to the valenceband (the energy bands in semiconductor mate-

rial), by recombining electron-hole pairs, an elec-tron gives up energy. In a LED, this energy is anemitted photon of light with a wavelength deter-mined by the band gap separating the two bands.

Emission occurs when current from the externalcircuit passes through the LED. With a photodiode,the opposite phenomenon occurs: light falling on

the diode creates current in the external circuit.

Absorbed photons excite electrons from thevalence band to the conduction band, a process

known as intrinsic absorption. The result is the cre-ation of an electron-hole pair. These carriers, underthe influence of the bias voltage applied to the

diode, drift through the material and induce acurrent in the external circuit. For each electron-hole

pair thus created, an electron is set flowing ascurrent in the external circuit. Several types of semi-conductor detectors can be used in fiber-opticsystems the pn photodiode, the pin photodiode,and the avalanche photodiode.

The pn Photodiode

The simplest device is the pn photodiode. (Referto Figure 2-16.) Two characteristics of this diode,however, make it unsuitable for most fiber-optic

applications.

First, because the depletion area is a relatively

small portion of the diodes total volume, many ofthe absorbed photons do not result in externalcurrent. The created hole and free electrons

recombine before they cause external current. Thereceived power must be fairly high to generateappreciable current.

Second, the slow tail response from slow diffusionmakes the diode too slow for medium- and high-speed applications. This slow response limitsoperations to the kilohertz range.

The pin Photodiode

The pin photodiodeis designed to overcome thedeficiencies of its pn counterpart. While the pindiode works like the pn diode, it has its peak sen-

sitivity to light signals at 1,000-2,000 nanometersin wavelength and can be used with LED sources

and medium- to high-loss fiber.

The name of the pin diode comes from the layeringof its materials: positive, intrinsic, negativepin.(Refer to Figure 2-17.) Care must be exercised in

2-13


n p

Figure 2-16PN Photodiode

i

n

p+

Figure 2-17PIN Photodiode


20/53

selecting the supplier of this important element ofthe fiber-optic system. It should be understood that

a tradeoff exists in arriving at the best pin photo-diode structure and balancing the opposing require-ments to achieve the best balance betweenefficiency and speed.

Avalanche Photodiode

The avalanche photodiode(APD) is morecomplex, consisting of more layers of silicon mate-rial than the pin photodiode. The APD, which wasdeveloped specifically for fiber-optic applications,

is efficient across a wider spectrum of light fre-quencies, suffers from less interference, and has afaster response time to signals than the pin photo-

diode. It is, however, more expensive as well.

NOISE

Noise (any electrical or optical energy apart from

the signal itself) is an ever-present phenomenonthat seriously limits the detectors operation. If thesignal is wanted energy, then noise is anythingelsethat is, unwanted energy.

Although noise can and does occur in every partof the system, it is of greatest concern in thereceiver input because the receiver works with

very weak signals that have been attenuated

during transmission. An optical signal that is tooweak cannot be distinguished from noise. To

detect such a signal, either the noise level must bereduced or the power level of the signal must beincreased.

An understanding of two types of noise, shot noiseand thermal noise, are important to the under-standing of fiber optics:

Shot Noise

Shot noise arises from the discrete nature of elec-

trons. Current is not a continuous, homogeneousflow. It is the flow of individual discrete electrons.

Remember that a photodiode works because anabsorbed photon creates an electron-hole pair

that sets an external electron flowing as current.It is a three-step sequence: photoelectron-holecarrierselectron. The arrival and absorption ofeach photon and the creation of carriers are partof a random process. It is not a perfect homoge-

neous stream, rather it is a series of discreteoccurrences. Therefore, the actual current fluctu-

ates as more or less electron holes are created inany given moment. Shot noise occurs even withoutlight falling on the detector.

Thermal Noise

Thermal noise, also called Johnson or Nyquist

noise, arises from fluctuations in the load resis-tance of the detector.

Thermal and shot noise exist in the receiver inde-

pendently of the arriving optical power. They resultfrom the very structure of matter. They can be min-imized by careful design of devices and circuits,

but they cannot be eliminated. For this reason thesignal must be appreciably larger than the noise inorder to be detected.

As a general rule, the optical signal should betwice the noise current in order to be detected.

SIGNAL-TO-NOISE RATIO

The signal-to-noise ratio (SNR) is a common way

of expressing the quality of signals in a system.SNR is simply the ratio of the average signalpower to the average noise power from all noisesources.

BIT-ERROR RATE

For digital systems, bit-error rate (BER) usuallyreplaces SNR as a measure of system quality.BER is the ratio of incorrectly transmitted bits to

correctly transmitted bits. A ratio of 10-9 meansthat one wrong bit is received for every one-billionbits transmitted.

DETECTOR CHARACTERISTICS

The characteristics of interest are those that relatemost directly to use in a fiber-optic system. Thesecharacteristics are:

Responsivity: The ratio of the diodes output

current to input optical power. It is expressedin amperes/watt (A/W).

Quantum Efficiency: The ratio of primary elec-tron-hole pairs (created by incident photons) to

2-14



21/53

the photons incident on the diode material).This deals with the fundamental efficiency of the

diode for converting photons into free electrons.

Dark Current: The thermally generated currentin a diode; it is the lowest level of thermal noise.

Minimum Detectable Power: The minimumpower detectable by the detector determined

the lowest level of incident optical power thatthe detector can handle.

Response Time: The time required for the pho-

todiode to respond to optical inputs andproduce external current. Usually expressedas a rise time and a fall time, measured in tens

of nanoseconds.

TRANSMITTERS AND RECEIVERS

BASIC TRANSMITTER CONCEPTS

The transmitter contains a driver and a source.(Refer to Figure 2-18.) The input to the driver is thesignal from the equipment being served. The

output from the driver is the current required tooperate the source.

Most electronic systems operate on standard,well-defined signal levels. Television video signals

use a 1 volt peak-to-peak level.

Digital systems use different standards, dependingon the type of logic circuits used in the system.

These logic circuits define the levels for the highsand lows that represent the 1s and 0s of digitaldata. Digital logic circuits, all further defined under

the Glossary in section 3, are:

Transistor-transistor logic (TTL) used in manyapplications.

Emitter-coupled logic (ECL), faster than TTLand not able to be mixed with TTL, it is usually

found in high-speed systems.

Complementary metal-oxide semiconductor(CMOS), which is rapidly becoming thereplacement for TTL because of its very lowpower consumption.

The drive circuits of the transmitter must acceptsignal input levels, then provide the output current

to drive the source. Characteristics for specifying atransmitter (or a receiver) are basically the same aswould apply for any electronic circuit. These include:

Power supply voltages

Storage and operating temperature ranges.

Required input and output voltage levels

(which indicate video, audio, TTL or ECL com-patibility).

Data rate/bandwidth.

Operating wavelength.

BASIC RECEIVER CONCEPTS

The receiver contains the detector, amplifier, andoutput circuit. (Refer to Figure 2-19) The amplifier

amplifies the attenuated signal from the detector.

The output circuit can perform many functions,such as:

Separation of the clock and data.

Pulse reshaping and retiming.

Level shifting to ensure compatibilityTTL,ECL, and so forthwith the external circuits.

Gain control to maintain constant levels inresponse to variations in received opticalpower and variations in receiver operation

from temperature or voltage changes.

Because the receiver deals with highly attenuatedlight signals, it can be considered the principalcomponent around which the design or selection

2-15


Driver

Source

Figure 2-18Basic Transmitter Block Diagram

Amplifier

Output Circuit

Detector

Figure 2-19Basic Receiver Block Diagram


22/53

of a fiber-optic system revolves. It is in the photo-detector and first stage of amplification that the

signal being transmitted is at its weakest and mostdistorted. It is reasonable to say that this is thecentral part of the link. Thus decisions affectingother parts of the link are made with the receiver in

mind. Decisions about the modulation of the trans-mitter are decided, at least in part, by the require-ments of the receiver.

Important receiver characteristics include:

Power supply voltages

Storage and operating temperature ranges.

Required input and output voltage levels(which indicate TTL or ECL compatibility).

Data rate/bandwidth.

Sensitivity.

Dynamic range.

Operating wavelength.

Sensitivityspecifies the weakest optical signal thatcan be received. The minimum signal that can bereceived depends on the noise floor of the receiver

front end.

Dynamic rangeis the difference between theminimum and maximum acceptable power levels.

The minimum level is set by the sensitivity and is

limited by the detector. The maximum level is setby either the detector or the amplifier. Power levels

above the maximum saturate the receiver or distortthe signal. The received optical power must bemaintained below this maximum.

AMPLIFIERS

The two most common designs found in fiber-optic

receivers are low-impedance amplifierand tran-simpedance amplifier. (See Figure 2-20.)

DUTYCYCLE IN THE RECEIVER

The reason for concern for duty cycle in the modu-lation codes is that some receiver designs put

restrictions on the duty cycle. A receiver distin-guishes between high and low pulses by main-

taining a reference threshold level. A signal levelabove the threshold is seen as a high or 1; a signallevel below the threshold is seen as a low or 0.Theshifting of threshold level would cause no problems

in an ideal, noiseless receiver. But receivers areneither perfect or noiseless. Signal levels not only

vary somewhat, but the signals also contain noise.

There are two ways to get around such errors. Thefirst is to maintain a duty cycle close to 50 percent.

Manchester and biphase-M codes, by definition,always have a 50 percent duty cycle, so theysatisfy the requirement. Their drawback is that they

require a channel bandwidth of twice the data rateand they also increase the complexity of the trans-mitter somewhat.

The second method of avoiding bit errors is todesign a receiver that maintains the thresholdwithout drift. The reference threshold is always

midway between high and low signal levels. Oneway to do this is by a dc-coupled receiver, which isdesigned to operate with arbitrary data streams.

The receiver is edge-sensing, meaning that it issensitive to changes in level and not to the levelsthemselves.This type of receiver reacts only to

pulse transitions.

TRANSCEIVERS AND REPEATERS

A transceiver is a transmitter and a receiver pack-aged together to permit both transmission andreceipt of signals from either station.

A repeater is a receiver driving a transmitter. It's

used to boost signals when the transmission dis-tance is so great that the signal will be too highly

attenuated before it reaches the receiver. Therepeater accepts the signal, amplifies and reshapesit, and feeds the rebuilt signal to a transmitter.

2-16


Low-Input-ImpedanceAmplifier

Low-Impedance Amplifier Receiver

High-Gain Amplifier

Transimpedance Amplifier Receiver

Figure 2-20Low- and Transimpedance Amplifiers


23/53

CONNECTORS AND SPLICES

The requirements for fiber-optic connection andwire connection are very different. In wiring, two

copper conductors can be joined directly bysolder or by connectors that have been crimped

or soldered to the wires. The purpose is to createcontact between the mated conductors to main-tain a path across the junction.

In fiber-optics, the key to interconnection isprecise alignment of the mated fiber cores (orspots in the case of a single-mode fibers) so that

nearly all of the light is coupled from one fiberacross the junction into the other fiber. Preciseand careful alignment is vital to the success of

system operation.

CONNECTOR REQUIREMENTS

Connectors provide the mechanical means for ter-

minating optical fibers to other fibers and to activedevices, thereby connecting transmitters, receivers,and cables into working links.

The primary task of the fiber optic connector is tominimize the optical loss across the interface of thecoupled fiber. This loss is expressed in decibels(dB). High-performance connectors are classified

as those with less than 1 dB of loss; medium perfor-mance is less than 2 dB. Losses occur from inexactmating of the fibers, and the surface condition of

the fiber ends. (See Figure 2-21.)

The second task of the connector is to provide

mechanical and environmental protection and sta-bility to the mated junction. Lastly, the connectordesign should permit rapid and uncomplicatedtermination of a cable in a field setting.

An ideal connector would encompass:

A fiber-alignment scheme yielding low loss. Physically small.

Rugged construction. Easily field terminated. Field repairable.

Good thermal characteristics. Offer excellent fiber/cable strain relief.

Accessory tooling to prepare fiber and cable. Factory terminated cable assemblies which

enable users to field connectorize or spliceassemblies using fusion or mechanical splices.

Be of moderate cost.

CAUSES OF LOSS IN AN INTERCONNECTION

Losses in fiber-optic interconnections are caused

by three factors: (1) Intrinsic, or fiber-relatedfactors caused by variations in the fiber itself.; (2)extrinsic, or connector-related factors contributedby the connector itself; or (3) system factors con-

tributed by the system itself.

In joining two fibers together it would be nice to

safely assume that the two are identical. However,the fact is that they usually are not. The fiber man-ufacturing process allows fibers to be made onlywithin certain tolerances.

Under section 3, Table G, Intrinsic Loss Factors,lists the most important variations in tolerances that

cause intrinsic loss, i.e., core diameter, claddingdiameter, numerical aperture mismatch, concen-tricity, ellipticity (or ovality) of core or cladding.

Connectors and splices contribute extrinsic loss tothe joint. The loss results from the difficulties inherentin manufacturing a connecting device to the exacting

tolerances that are required. The four main causes ofloss that a connector or splice must control are:

Lateral displacement: A connector shouldalign the fibers on their center axes. When onefibers axis does not coincide with that of theother, loss occurs.

2-17


Figure 2-21Diameter Mismatch of Connectors

Core

Core 1

Core 2

Cladding

Cladding

Ellipticity (Ovality)

Cladding Diameter Mismatch

Core Diameter Mismatch

Concentricity


24/53

End separation: Two fibers separated by asmall gap will suffer loss.

Angular misalignment

Surface roughness.

Again, see Figure 2-21 on the previous page.When two fibers are not perfectly aligned on their

center axes, lateral displacement loss occurs evenif there is no intrinsic variation in the fiber.

First, the fiber ends must be optically square and

smooth, and second the end-to-end presentationof both fibers must align and the gap (air space)be made minimal. In the case of single-mode con-

nectors, the fiber ends may come into contact toreduce the reflective losses.

Two fibers separated by a small gap experienceend-separation loss of two types. First is a Fresnelreflection loss caused by the difference in refrac-

tive indices of the two fibers and the interveninggap, which is usually air. The second type of lossfor multimode fibers results from the failure ofhigh-order modes to cross the gap and enter the

core of the second fiber.

Either of these conditions will contribute to loss,the result being dependent on the numerical aper-

ture (NA) of the fiber.

A gap between a transmitting and a receiving fiber

will also introduce loss because the air betweenthe fibers is of a different refractive index than thecore of the fibers. With air between the fibers, theFresnel loss would be 0.4 dB. This can be reduced

by immersing the junction in a fluid of matchingliquid, typically with a refraction index the sameas that of the core. Some connectors use this

feature, but at the risk of fluid depletion and pos-sible introduction of contaminants.

The ends of mated fibers should be perpendicular

to the fiber axes and perpendicular to each other.In order to ensure this, fiber ends are made squareand smooth by one of two methods. These are the

lap-and-polish (grind) method and the scribe-and-break (cleave) method. The lap-and-grind method

involves the use of a positioning fixture andgrinding/lapping compounds.

Once the ends are square and smooth, the connectordesign must address alignment parameters to ensure

lowest loss. In particular, the connectors must mini-mize fiber lateral offset and angular misalignment.

Finally, the fiber face must be smooth and free ofdefects such as hackles, burrs, and fractures.

Irregularities from a rough surface disrupt the geo-metrical patterns of light rays and deflect them so

they will not enter the second fiber, thus causingsurface finish loss.

System-related factors can also contribute to lossat a fiber-to-fiber joint. Refer to page 2-6, wherethe subject of dispersion is discussed, and specif-ically describes how modal conditions in a fiber

change with length until the fiber reaches equilib-rium mode distribution (EMD).

Initially, a fiber may be over filled or fully filled withlight being carried both in the cladding and inhigh-order modes. Over distance, these modes

will be stripped away. At EMD, a graded-indexfiber has a reduced NA and a reduced active areaof the core carrying the light.

Consider a connection close to the source. Thefiber on the transmitting side of the connectionmay be over filled. Much of the light in the

cladding and high-order modes will not enter thesecond fiber, although it was present at the junc-tion. This same light, however, would not havebeen present in the fiber at EMD, so it would also

not have been lost at the interconnection point.

Next consider the receiving side of the fiber. Some

of the light will spill over the junction into claddingand high-order modes. If the power from a shortlength of fiber were to be measured, these modeswould still be present. But these modes will be lost

over distance, so their presence is misleading.

Similar effects will be seen if the connection point

is far from the source where the fiber has reachedEMD. Since the active area of a graded-index fiberhas been reduced, lateral misalignment will notaffect loss as much, particularly if the receiving

fiber is short. Again, light will couple into claddingand high-order modes. These modes will be lost ina long receiving fiber.

Thus, the performance of a connector depends on

modal conditions and the connectors position inthe system. In evaluating a fiber-optic connectoror splice, we must know conditions on both thelaunch (transmitter) side and the receive (receiver)side of the connection.

2-18



25/53

Four different conditions exist:

Short launch, short receive.

Short launch, long receive.

Long launch, short receive.

Long launch, long receive.

LOSS IN SINGLE-MODE FIBERS

It is important to note that connectors and splices forsingle-mode fibers must also provide a high degreeof alignment. In many cases, the percentage of mis-

alignment permitted for a single-mode connection isgreater than for its multimode counterpart. Becauseof the small size of the fiber core, however, the

actual dimensional tolerances for the connector orsplice remain as tight or tighter.

ALIGNMENT MECHANISMS ANDSPLICE EXAMPLES

Many different mechanisms have been used toachieve the high degree of alignment that isrequired in a connector or splice. Splicing is the

name of the process whereby two fibers or cables

are joined together. Fiber splicing consists of:preparation of the fiber; cleaving the fiber; inspec-

tion of the cleave; placing of the fibers in an align-ment fixture; alignment or tuning of fibers; bondsplice; inspection and testing; and enclosing ofthe splice for protection.

Basically, there are two types of splices: fusionand mechanical.

FUSION SPLICES

The fusion splice is accomplished by applyinglocalized heating at the interface between twobutted, prealigned fiber ends, causing the fibers

to soften and fuse together to form a continuousglass strand. This system offers the lowest light

loss and the highest reliability. Loss should be at.5 dB/splice or less.

Specifically, the fusion splice consists of:

Joining glass fibers by melting them togetherusing an electric arc.

Precision controlled for fiber uniformity.

Permanent, highly reliable, low in cost.

Average of 50 splices can be done per day inone location by a single team of two persons.

Typically 0.1 to 0.3 dB loss per splice.

A fusion-splice joint can maintain a breaking strainof more than one percent. This means that such

splices can be used when manufacturing fiber-optic cable if long, continuous cables of tens ofkilometers are required.

The down side of this method is that training isrequired before using the expensive equipment

that effects the fusion splice. Depending on thecomplexity of the installation, this may not be thefirst choice.

The fusion-splice process employed can varydepending on the type of splicer used. The twomost common types are the local injection detection(LID) splicerand the manual splicer. Both splicersuse electrodes to melt the fiber ends together.

The LID Splicer

The LID spliceror automatic splicer, is a processthat employs microbending techniques to launch

light into the fiber before the fiber end. On theopposite fiber to be spliced a microbend is againused, but this time with a detector to remove the

launched light. This allows the processor in thesplicer to align the fiber to where the greatestoptical power level is achieved.

The process for this splicing is positioning the fiberin clamps and alignment fixtures. By activating theautomatic alignment function, the splicer runs

though various X, Y, and Z alignments for opti-mizing the transmission through the two fiber ends.

When this is accomplished, the splicer indicatesmaximum alignments and the splicer operator thenfuses the fibers by activating electrodes.

2-19


SECTION 2 FIBER OPTIC BASICS


26/53

The Manual Splicer

A manual splicerusually has two alignment fixtures,each located on one side of the splicer permitting

manual aligning of fiber end through X, Y, and Z axes.

The splicer having prepared each fiber for splicingthen places the fibers in clamps located on eachside of the electrodes. The clamp and alignment

fixtures are then manually manipulated while thesplicer views the process through a microscope.In this process the splicer can inspect the fiberends and the alignment process.

The manual fusion splicer is less expensive thanthe local injection detection splicer and is good for

making multimode splices. Because this unitaligns the fibers on the outer diameter of the fibers,losses can be slightly higher than a LID set which

optimizes the fiber cores.

It should be noted that because all f ibers are not

identical, a good fusion splicer should be easilyadjustable to change arc duration and current tothe electrodes. The reason is that different fiberscan melt or fuse at different temperatures and

require longer or slower fusion arcs.

Further, when using LID systems, the techniqueallows for optimum core alignment. However, the

measurements obtained from this technique may

not match the OTDR measurements which wouldbe optimized using the same wavelength that the

system would operate at.

MECHANICAL SPLICE

Mechanical splices are the most straightforward.The installer merely terminates the two ends of thecable that are to be joined and then connects

them with an inexpensive barrel splice.

This method is fine for short-haul systems, butintroduces light loss of up to 4 dB/splice that may

degrade a system that operates over a distancegreater than two kilometers. It consists of:

Fibers joined by a glass capillary.

Splice is permanent, with good reliability and

low loss.

Average of 50 splices per day in one location.

Typically 0.1 to 1dB loss per splice (at 850 or1300 nm).

Can be reusable.

Mechanical splicing methods include rotary,central glass alignment guide (or four-rod), andelastomeric.

The Rotary Splice

The rotarysplice (see Figure 2-22 ) is a newermethod of splicing optical fibers. The rotary is both

a connector and a splice as it does have thecapability to be mated and unmated like connec-tors, yet has the low attenuation features of anoptical splice. Like optical connectors, this splice

takes longer to terminate, requires more compo-nents, and has a higher component failure rateprior to testing.

Central Glass Alignment Guide Splice

The central glass alignment guide spliceuses fourprecision glass rods to precisely align optical fibers.The rods are fused together creating an inner hollowcore. At each end of the splice, the rods are bent ata slight angle allowing the fibers to orient themselves

in the uppermost V groove of the rods. By positioningthe fiber where the ends will be in the middle of thesplice, the fibers can be precisely rotated to allow

for the lowest attenuation.

2-20


Spring Retainer

LG Fiber Alignment Sleeve LG Fiber

Compression Spring

Ferrule

Figure 2-22Rotary Splice



27/53

With the use of splice holders, this type of splice canbe used for temporary splices in both lab and field

applications. By using a splice holder, the splice iseasier to work with and has a substantially lower

discard rate due to its alignment rod technique.

For permanent installations, the hollow section withthe rods is filled with UV fluid. After aligning thescribed fibers, the splice is cured in minutes by

using a UV lamp. Like all good splices, theprocess requires a good end face to maintain lowattenuation. The advantage of this type of spliceare versatility for field and lab applications and low

tooling costs.

Elastomeric Splice

The elastomericsplice (Figure 2-23) is made froma plastic (elastic) material formed into a mold. The

mold allows for a hole to be made. The elas-tomeric material is flexible enough so the fiberscan be positioned and firm enough so the fibers

are retained during handling and splicing withoutthe need for positioning equipment.

Because the fibers are mated into the same mold,

alignment can be maintained with low attenuation.The fibers can be tuned for low attenuation if careis taken in removing the fibers prior to tuning. Likethe central glass alignment method, the elastomeric

method uses matching fluids or UV fluids dependingon the application. The need for a good scribed

optical fiber will allow for low attenuation measure-ments. A typical elastomeric splice will introduce

light loss of less than 1 dB/splice.

FIBER PREPARATION

Proper preparation of the fiber end face is critical

to any fiber-optic connection. The two main fea-tures to be checked for proper preparation areperpendicularity and end finish.

The end face ideally should be perfectly square tothe fiber and practically should be within one or twodegrees of perpendicular. Any divergence beyond

two degrees increases loss unacceptably. The fiberface should have a smooth, mirrorlike finish freefrom blemishes, hackles, burrs, and other defects.

The two most common methods used to producecorrect end finishes are the cleaving(or scribe-

and-break) method and the polishmethod. Thefirst is used with splices and the second is morecommonly used with connectors.

Whichever method is used, it is necessary to preparea fiber for splicing. To do this the protective jacketsand buffers must be removed to allow access to theoptical fiber. The outer and the inner jackets are

2-21


Fiber

End Guide

End Guide

Fiber

ElastomerInserts

GlassSleeve

Outer

Cylindrical

Sleeve

V-GrooveTempered

Entrance Hole

Insert Parts

Figure 23Elastomeric Splice



28/53

removed, exposing the Kevlar strength member, thebuffer tube, and the fiber. The fiber still has the pro-

tective coatings which will also have to be removed.

Standard cable strippers can be used to removethe outer jacketing. The amount of Kevlar removed

can vary depending upon the design of thestrength member of the cable. If the cable doesnot incorporate a strength member, the Kevlar can

be used as such.

The buffer tubes, like the outer jackets, can beremoved by mechanical stripping tools with the

operator taking care not to kink or damage theinternal coated fibers.

Once the coated fiber is exposed, the splicer mustremove the protective coatings to start the actualfiber splicing. Most coated fibers can be stripped

using mechanical or chemical methods. Thesplicer should also take care to use tools or proce-dures that will not damage the fibers.

After the coating is removed, the splicer shouldclean the fiber with Isopropyl alcohol to assure thatthe fiber is clean. Contaminants on fiber can cause

the fiber to misalign itself in the alignment fixture.

Cleaving

This is a process which allows the operator tobreak or scribe the fiber with a 90 degree end

face perpendicular to the axis of the fiber with l

Documents

Optic Fiber Basics